This project aims to classify the emotion on a person's face into one of seven categories, using deep convolutional neural networks. The model is trained on the FER-2013 dataset.
To install the required packages, run pip install -r requirements.txt.
1.Python
2.Convolution Neural Network(CNN)
3.Open CV
4.Data Augmentation
First, clone the repository and enter the folder src
Download the FER-2013 dataset from here and unzip it inside the src folder. This will create the folder data.
I had added the dataset into gitignore as it is a very big file.
If you want to train this model, use:
cd src
python emotions.py --run train
If you want to view the predictions without training again, you can download the pre-trained model from here.
If you want to use the web camera or give input as a video to detect emotions just run- cd src python emotions.py --run test with USE_WEBCAM as True or False
If you want to detect emotions in a picture just run- cd src python emotions.py --run picture
With a simple 4-layer CNN, the test accuracy reached 63% in 50 epochs.
The original FER2013 dataset in Kaggle is available as a single csv file.
Model Summary--
